Climate change can increase the risk of conditions that exceed human thermoregulatory capacity1,2,3,4,5,6. Although numerous studies report increased mortality associated with extreme heat events1,2,3,4,5,6,7, quantifying the global risk of heat-related mortality remains challenging due to a lack of comparable data on heat-related deaths2,3,4,5. Here we conducted a global analysis of documented lethal heat events to identify the climatic conditions associated with human death and then quantified the current and projected occurrence of such deadly climatic conditions worldwide. We reviewed papers published between 1980 and 2014, and found 783 cases of excess human mortality associated with heat from 164 cities in 36 countries. Based on the climatic conditions of those lethal heat events, we identified a global threshold beyond which daily mean surface air temperature and relative humidity become deadly. Around 30% of the world’s population is currently exposed to climatic conditions exceeding this deadly threshold for at least 20 days a year. By 2100, this percentage is projected to increase to ∼48% under a scenario with drastic reductions of greenhouse gas emissions and ∼74% under a scenario of growing emissions. An increasing threat to human life from excess heat now seems almost inevitable, but will be greatly aggravated if greenhouse gases are not considerably reduced.
We thank the Gridded Human Population of the World Database and the National Center for Environmental Prediction and Department of Defense reanalysis database for making their data openly available and B. Jones for sharing human population projections. We acknowledge the World Climate Research Programme’s Working Group on Coupled Modelling, which is responsible for CMIP5, and thank the climate modelling groups (listed in Supplementary Table 1) for producing and making available their model outputs. We also thank D. Schanzenbach, S. Cleveland and R. Merrill from the University of Hawai’i Super Computer Facility for allowing access to computing facilities and Hawai’i SeaGrant for providing funds to acquire some of the computers used in these analyses. Q. Chen, A. Smith, C. Dau, R. Fang and S. Seneviratne provided valuable contributions to the paper. The opinions or assertions contained herein are the private views of the authors and are not to be construed as official or as reflecting the views of the Army or the Department of Defense. We thank R. Carmichael, M. Deaton, D. Johnson and M. Smith in ESRI’s Applications Prototype Lab for the creation of the online mapping application. This paper was developed as part of the graduate course on ‘Methods for Large-Scale Analyses’ in the Department of Geography, University of Hawai’i at Mānoa.